BlackForestLabs Announces and Upcoming Open-Source Release of FLUX 3 Video Model
歸藏的AI工具箱 · wechat · 2026-07-23
Black Forest Labs has officially released the FLUX 3 video generation model. It boasts a comprehensive feature set, including text-to-video, image-to-video, reference-clip based generation, audio/video extension, keyframe-based generation, and multi-shot sequence linking, alongside image and animation design outputs.
The developer announced that the FLUX 3 Dev model will be open-sourced. While early access requires application currently, API access and the open-source version are expected to roll out in the coming weeks.
Related event: Black Forest Labs Launches FLUX 3: A Unified Multimodal Foundation Model(28 posts)→
More from Multimodal
- Scribbly AI art style shown across multiple generated image examples — OVolosin82152 · 2026-07-24
- FLUX 3 expands into one multimodal backbone for image, video, audio and action — pmttyji · 2026-07-24
- Flux 3 X Mimic debuts as a next-generation video-action model — kensai · 2026-07-24
- Flux 3 is being tested on perfectly synced split-screen video generation — umesh_ai · 2026-07-24
- Flux 3 appears to be a capability family, not a separate "Klein" model — carrot_2333 · 2026-07-24
- A user asks why 4x UltraSharp upscaling is producing strange image artifacts — MeDotEE · 2026-07-24